Skip to content
Pylon Digital

AI for law firms · Guide

How to automate law firm client intake, and what must stay human

To automate law firm client intake, let software handle the repeatable steps: reading website and email enquiries, asking for missing details, preparing conflict-check prompts and proposing meeting times. A lawyer still decides every conflict, whether the firm acts, and anything that amounts to advice. This guide sets out the steps, the privacy notice and the controls.

Published
Last reviewed
Reading time
7 min read

What does it mean to automate law firm client intake?

Automating client intake means software handles the administrative steps between a first enquiry and a first meeting, while lawyers keep every decision. The software reads the enquiry, drafts a reply asking for missing details, checks names against your records and proposes meeting times. A lawyer decides whether there is a conflict, whether the firm acts, and what the enquirer is told.

Most firms already have these steps. They are just done by hand, often by whoever is free when the enquiry lands. The table shows where the line between software and people usually sits.

Intake stepWhat automation doesWhat a person does
Enquiry arrives by web form or emailLogs it, reads it and drafts a reply asking for missing detailsApproves the reply before it is sent
QualifyingCompares the enquiry with your written criteria: practice area, jurisdiction, matter typeDecides whether the firm will act
Collecting detailsCaptures names, contact details, other parties and key dates in set fieldsChecks anything unclear
Conflict-check promptSearches the practice-management system for the client, other parties and related names, then lists possible matchesA lawyer decides whether a conflict exists
BookingProposes times from the right lawyer’s calendarConfirms the booking
EngagementPrepares the costs disclosure and engagement letter from your templatesReviews, signs and sends

How do you automate law firm client intake, step by step?

Start with your own rules, not the software. A firm that writes down how it qualifies enquiries, what it collects and who decides will find automation straightforward; a firm that skips this step automates its inconsistencies. These seven steps suit a 10–100 person firm, and they can be piloted in one practice area first.

  1. Write your qualifying rules. List the practice areas and jurisdictions you act in, the matter types you decline, and the details a lawyer needs before a first meeting. A commercial team, for example, might need the other party’s full name and ABN, plus any contract or notice dates.
  2. Write the collection notice. Draft the privacy wording that appears wherever details are collected (see below), and have it reviewed.
  3. Map the fields to your system. Decide where each detail lands in Actionstep, LEAP or Clio, such as a prospective-matter record with the other party as a linked contact. Actionstep and Clio both publish developer APIs (Actionstep, Clio); the connection method for your system is confirmed before any build.
  4. Set the conflict-check prompt. Specify which names are searched, including related companies, trading names and former names, and who receives the result.
  5. Connect booking to calendars. Route each practice area to the lawyers who take that work, and only offer times once the conflict check has cleared.
  6. Set approval points and logging. Decide which messages need approval. With the Client Intake Agent, by Pylon Digital, that is every message leaving the firm, and every action is written to an audit log.
  7. Pilot and review. Run one channel or practice area for an agreed trial period, with a lawyer reviewing every record, before widening it.

Where does the conflict check sit, and who decides?

Software can run the search; a lawyer must make the decision. The Australian Solicitors’ Conduct Rules require solicitors and law practices to avoid conflicts between the duties owed to current and former clients (rule 10.1) and between two or more current clients (rule 11.1). Whether a match is a real conflict, and whether informed consent or an effective information barrier changes the answer, is a professional judgement.

Good conflict-check prompts are built to surface possible matches, not to clear them. That means searching the prospective client, every other party and related entities such as a company’s directors or a trading name, and treating near matches (“J. Nguyen” and “Jenny Nguyen Pty Ltd”) as matches until someone has looked. The result should show the reviewing lawyer the search terms used and every hit, so the decision can be checked later.

Sequence matters too. Collect only what the check needs, such as names, the other side and the matter type, until it clears. If the firm cannot act, it should not be holding a prospective client’s full account of their dispute.

What must stay with a person?

Anything that involves judgement, advice or a decision about someone’s position stays human. Automation is useful precisely because it does not make these calls: it prepares the record so a lawyer can make them faster.

  • Deciding to act or decline. A lawyer decides, and a person approves the wording of any decline.
  • Conflicts. The lawyer decides; the software reports what it found.
  • Anything that reads as advice. Intake replies should not comment on prospects, time limits or likely costs. If an enquiry mentions an accident date or a deadline, the software flags it for urgent review instead of commenting.
  • Costs. Costs disclosure is a legal obligation of the law practice (in Uniform Law states, section 174 of the Legal Profession Uniform Law). Software can prepare the document from your template, but a lawyer checks the estimate.
  • Urgent or sensitive situations. Enquiries that mention family violence, an arrest, an imminent hearing or someone at risk should skip the queue and reach a person straight away.

There is a privacy reason as well. From 10 December 2026, an organisation covered by the Privacy Act that arranges for a computer program to make, or do a thing substantially and directly related to making, a decision that could reasonably be expected to significantly affect an individual’s rights or interests must describe this in its privacy policy (OAIC guidance on APP 1). Whether automated qualifying of enquiries falls within that is a question for your privacy adviser before go-live.

What privacy notice do you need when you collect enquiry details?

If your firm is covered by the Privacy Act, APP 5 requires reasonable steps to tell people certain things at or before the time you collect their personal information, or as soon as practicable afterwards. The Act covers organisations with annual turnover above $3 million and some smaller businesses (OAIC). For intake, the notice belongs beside the form and in the first reply.

The APP 5 guidelines set out the matters to cover, which include who you are and how to contact you, why you collect the information, the main consequences if it is not provided, who you usually disclose it to, how people can access and correct it or complain (through your privacy policy), and whether it is likely to go overseas. The OAIC’s guidance on commercially available AI products adds that public-facing AI tools such as chatbots should be clearly identified as such.

Illustrative wording, for your firm to adapt and have reviewed:

We collect these details to respond to your enquiry, check for conflicts of interest and arrange a meeting. An AI system helps our staff sort and summarise enquiries, and a lawyer reviews each one. If you do not provide these details, we may not be able to help. Our privacy policy explains how to access or correct your information, how to complain, and whether we send information overseas.

Keep the form narrow. Health information is sensitive information, and APP 3 generally requires consent before it is collected. A personal injury enquiry form can ask for the date and type of incident without asking for a medical history.

What does automated intake look like in practice?

A worked scenario makes the division of labour clear. It is illustrative, not a client result.

A 25-person firm with commercial and property teams runs Actionstep. At 9.40 pm an enquiry arrives through the website about a retail lease dispute. By morning, the intake software has created a prospective-matter record with the tenant, the landlord and the landlord’s parent company as linked contacts, and searched all three. It finds that the parent company is a former client and flags the match. A partner reviews the old matter, decides the firm can act, and records the reason. The software then drafts a reply offering three times from the property partner’s calendar, which the practice manager approves. Every step, including the search terms used, sits in the audit log.

How does the Client Intake Agent, by Pylon Digital, fit?

The Client Intake Agent, by Pylon Digital covers the core of this guide: it qualifies website and email enquiries, collects details, runs conflict-check prompts and books meetings. It is one of our AI agents for professional firms, built with the same guardrails: human approval before anything leaves the firm, a full audit log, and client data stored in fully GDPR-compliant data centres.

If you are still deciding whether AI belongs in your firm at all, start with is AI safe for Australian law firms. Our page for law firms shows how intake fits alongside reporting and drafting.

This is general information, not legal advice.

Questions

Frequently asked questions

Can AI run conflict checks for a law firm?

AI can run the search, but a lawyer must make the decision. Intake software can search your practice-management system for the prospective client, the other parties and related entities, then list every possible match with the search terms it used. Whether a match is a real conflict under the Australian Solicitors' Conduct Rules, and whether consent or an information barrier could resolve it, is a professional judgement that stays with a lawyer.

Do we have to tell enquirers that AI is involved?

You should say so plainly. If your firm is covered by the Privacy Act, APP 5 requires a collection notice covering matters such as why you collect the details and who you usually disclose them to. The OAIC's guidance on commercially available AI products also says public-facing AI tools should be clearly identified as such. One plain sentence in the notice beside the form, repeated in the first reply, does the job.

Which practice-management systems can automated intake connect to?

The Client Intake Agent, by Pylon Digital, works with Actionstep, LEAP and Clio, and with Microsoft 365 for email and calendars. Actionstep and Clio both publish developer APIs. The right connection method depends on your system, version and licence, so we confirm it in the free 45-minute discovery call before anything is built or any data is connected.

What happens to enquiries the firm declines?

A lawyer makes the decision, and a person approves the decline message before it is sent. The message should not explain the law or mention time limits, because that can read as advice to someone who is not your client. The enquiry, the conflict search and the decision stay in the audit log, and your records policy sets how long the enquiry is kept.

Where is intake data hosted?

It depends on every tool in the chain. The web form, the email system, the practice-management system and any AI service each hold a copy somewhere, so ask each vendor for the region, retention period and training settings. The Client Intake Agent, by Pylon Digital, stores its data in fully GDPR-compliant data centres. Our security and data residency page explains the detail.

Secure by design. Set up correctly. Fully managed.

Talk to us before you commit to anything

Start with a free 45-minute discovery call. We look at your systems and priorities, then recommend a first step with a fixed scope, or tell you if we are not the right fit.

Book a free 45-minute discovery call